Related to: Building rationalist communities, Lessons from Latter-day Saints, Holy Books (Or Rationalist Sequences) Don't Implement Themselves, Designing rationalist projects, Community roles: teachers and auxiliaries.

The ultimate question I’m trying to answer is: what should be the roles in a rationalist community?           

In the previous post and this one, I outline the roles in Latter-day Saint communities along with some implications for rationalist communities. In the following posts, I will elaborate the analysis of implications.

Good sets of rules resemble programs; once developed, they can be made to run everywhere, with local modifications.[1]


In the previous post, I discussed the role of teachers and auxiliaries. Teachers are the ones who speak, teach and lead discussions in Sunday church meetings. Auxiliaries are responsible for ensuring the well-being of the various segments of the congregation – men, women, and children.

In this post, I will discuss committees and leadership.

Committees are in charge of specific important tasks – planning activities, giving service, and so on. The general leadership appoints people to each of the positions, provides personal counseling, and plans the curriculum.

In smaller groups, the leaders and more-committed members often wear multiple hats.


In a Latter-day Saint congregation, where 100 to 150 people come to church every week, there are a lot of committees. Committees are made up of male and female co-chairs and as many members as are necessary and available.

Fellowship Committee. This exists to help church members who move in become integrated into the ward. They follow up with leadership to make sure move-ins meet with church leadership, receive a calling (=responsibility), get home or visiting teachers, get a home or visiting teaching assignment, and so forth.

In family wards, they make sure that half the men’s organization comes over on Sunday after church to load (or unload) the U-Haul. And that the women’s organization brings over desserts, food, etc at some later time.

Missionary Committee. This helps new people start coming to meetings. For example, setting an example by inviting friends, encouraging people to invite their friends, thanking people for bringing their friends, befriending new people when they show up, making sure meeting topics are understandable to new people, asking last week’s newcomer “Hey, are you coming tomorrow?” in the right way, and making sure that they are integrating into the social structure, similar to the fellowship committee.

Mormons have a checklist for this one. If they’re a new church member, they should have a friend, a responsibility, and be “nourished by the good word of God,” ie, they should be applying the principles they learn at church into their lives. Check check and check means we’ve done our job.

Family Home Evening. This used to be my responsibility. Basically my co-chair and I would organize fun activities every Monday evening. For example: standard party games (Mafia, Psychiatrist, Pictionary), volleyball, a charity date auction. The ideal activities involved breaking up into small groups and competing on a task: a scavenger hunt, a marshmallow-and-toothpick bridge-building challenge.

Service, Temple, and Family History committees. These three committees are “outward facing” and merit mention. Service will organize days to clean a local park (or similar group task). Temple will organize trips to special buildings called temples. Family History will lend assistance in researching one’s genealogy. Basically any church member can do these things; it requires no specialized knowledge.

Equivalent “committees” for Less Wrong could support, say, cryonics and Singularity, existential threat research. Of course, the tasks would have to be designed to require less specialized knowledge, similar to perhaps open-source hacking storms (a bunch of people spend 24 hours together writing code).

Housing. This one is simple but powerful. And it isn’t even a committee, it’s just one person who maintains a Google Doc of people looking for a place to live and people looking for roommates.

Consider: the church population in my church unit is entirely composed of young single adults. Over half of the church members who don’t live with their parents live with each other. This, predictably, helps to build friendships. It probably increases the regularity of attending church meetings.


As you might have guessed by now, one main function of the general congregation-level leadership is to make sure all of these gears are grinding smoothly.

Congregations are led by a bishop, who is chosen every five years from among the congregation. They are not paid – nor is anyone else in this structure. The bishop chooses two counselors, a secretary, and various clerks to keep the records and the books. He also selects the heads of each organization: men’s organization, women’s organization, teaching organization, committees.

Together with the bishop and his counselors, they choose which people should fulfill which roles.

A key concept in Latter-day Saint leadership is that of stewardship. As the name implies, the leaders have a spiritual responsibility over each member of their congregation. People have their agency, the freedom to choose, but the leaders have a charge to help them not go astray – ie, to keep coming to church, continue living the lifestyle. This responsibility is taken pretty seriously.

The main unique role of the bishop is to provide personal counseling services for the congregation. It’s understood that if you’re struggling with addictions, having marital difficulties, or other personal problems, you should go talk to the bishop. Even when that’s not the case, the system is set up so them you have some kind of interview with the bishop once or twice a year.

I’ve had good experiences with this, just having designated time to talk privately with an older and wiser person. Think of it as a universal free counseling service.

Another role is to effectively oversee the workings of the congregation. As with most leaders, bishops consciously and unconsciously shape the culture.

If they bring or invite friends to church, others will too. If they express frustration at the eccentricities of newcomers, others will too. If they don’t pay attention to new members and whether they are getting socially integrated into the church unit, others won’t either.

If the bishop and his counselors create a high meeting quality, people are far more likely to bring their friends.

There are various levels at which attrition operates; of which bishops and other leaders are well aware. These roles are also tools to help members not go astray.

The first level of attrition is that people stop coming because they don’t have any friends. For church newcomers, if this is the case, will generally happen in the first couple months after they start coming.

Solution for bishops: seriously ask specific congregation members to befriend them. Have the men’s and women’s organizations assign them high-quality home and visiting teachers.

The second level is when people haven’t really applied church principles into their everyday lives. They might come for the friends, and they feel happy when they come, but they have no deep investment in the lifestyle.

Something will come up, it always does. Maybe they will get married to an uninterested spouse, maybe they’re tired of constant snide remarks from disapproving family, maybe their new job has them work on Sunday, maybe a new set of friends who just have other things to do.

And so they stop coming.

Solution for bishops: ask them to accept a calling that gives them definite responsibilities and a place within a tightly-knit auxiliary or committee – say, Sunday School teacher for the 10 and 11-years olds.

The third level is a conscious, seriously considered decision by a longtime member to withdraw.

A lot of the time, maybe thirty to fifty percent, this is becoming "offended" and is due to decisions made or not made by the bishop. Perhaps he told them to do something they didn’t like, or there was a dispute and he took the other side, or they feel ignored and unappreciated, or they just don’t like his style of running the congregation.[2]

A good bishop is key to a healthy congregation.

[1] Paul Graham: “The other way makers learn is from examples…hackers, likewise, can learn to program by looking at good programs-- not just at what they do, but the source code too. A McDonald's franchise is controlled by rules so precise that it is practically a piece of software. Write once, run everywhere.”

[2] Or maybe the bishop does do something illegitimate. In one very troubled church unit I know of, three successive bishops were skimming tithing funds. (They are supposed to be unpaid.) They were caught and excommunicated, but all the old church members in the area stopped coming and the church unit still hasn’t recovered, ten to fifteen years later. But this kind of blatant misconduct is in, my experience, extremely rare.

New Comment
14 comments, sorted by Click to highlight new comments since:

Good sets of rules resemble programs; once written, they can be run everywhere.

Are you a programmer?

Paul Graham, who wrote something very similar which is referenced in a footnote, most certainly is.

This jumped out at me as well. Something like "once developed, they can be made to run everywhere" is probably a more intuitively reasonable wording that communicates the same intended message.

Point taken, thanks.

If you read pg's essay, his quote - within context - expresses something much more modest than what calcsam is saying within their own context.

There is something ironic, to me, in the bald assertion that "roles and rules" are like programs; it is a topic of hot debate within the software engineering community whether "methodologies", that is, programs of precisely this sort governing the creation of software by humans, can in fact be written once and run everywhere; they have been the focus of the discipline for over forty years of its history, but the consensus is, I think, tending away from this belief. (Here is a representative article.)

methodologies are a different kettle of fish to something like the rules of running a McD's.

The former tries to create rules around an essentially creative process. The latter covers something that really is not creative. Big Macs and clean floors are made the same way the world around - so a set of rules can in fact cover just about every likely eventuality in a regular McD's store.

Not so methodologies which attempt to place a "do it the same every time" framework (learned from such systems as franchise burger joints) to a process that, by it's nature must be different every time - and also involve creative scope that a set of rigid rules just can't handle.

My point? Methodologies really can't be written "like software" (at least not yet), but franchise-operating rules can.

I'm still struggling to explain why this is (to people who ask) without devolving into a "different magisteria" kind of explanation. :P

Yes to the above, for the most part.

The irony is that "good sets of rules resemble programs" is an intuition which may appeal most to people with limited understanding and experience of what it takes to create a useful and robust program.

In particular, I strongly doubt that there will be anything McDonalds'-like about a successful effort to raise the rationality waterline.

The further irony: we know that there is a set of rules for McDonald's which has program-like precision. However, we do not have much evidence that this program is followed with computer-like compliance; and in fact whenever I actually enter a McDonald's restaurant (which I grant is very seldom, but still) I find blatant evidence of non-compliance: dirty floors, french fries cold and slumped because they've sat too long before being served, burgers in similar condition...

More generally the existence of "work to rule" strikes suggests that, even when workers in a given context are popularly conceived to follow program-like rules with computer-like obedience, this is rarely in fact the case and these rules only appear to be effective precisely because workers in fact exert some significant degree of autonomy and judgement in carrying out those rules.

TL;DR: the more you know about either programming or McDonalds, the less grounds you have to think that "good sets of rules resemble programs".

In particular, I strongly doubt that there will be anything McDonalds'-like about a successful effort to raise the rationality waterline.

Yes I agree. Changing people's thinking falls into the "creative process" box for me.

However, we do not have much evidence that this program is followed with computer-like compliance

Good point. I guess the rules are "more like guidelines" - but when they're followed well they lead to a much higher chance of a successful McD's franchise... (though it'd be interesting to see if that pans out in a study).

Perhaps by analogy to territorial colonization? Programmers are like prospectors, trying to find shiny, expensive spots amid a vast desert of potential code sequences, the overwhelming majority of which are garbage. Once a section of idea-space is found to be valuable and habitable, franchise-builders are like early law enforcement: the law doesn't have to be perfect, there's no motherlode to aim for. It just has to lay down some standards the people are willing to accept, and then enforce them consistently enough that society can continue to stabilize.

A law like "don't murder people for pocket change" that people will accept in New York, they'll probably accept in Santa Fe too. But a rock that had gold under it in Santa Fe won't still have gold under it after you carry it back to New York.

I'm not sure you get the issue that is the problem.

think about it by comparing to another creative process: painting art.

What "methodology" can be put in place to always ensure that your art is exceptional? Sure - you can find techniques that have worked in the past... but if you enforce only one way to paint, then you'll destroy all creativity... which is kinda the point with art.

For programming - creativity in solving problems is the key. Every problem is different - and sure, there are techniques that have been useful in the past - but if you enforce them, rigidly - it stops a new. creative breakthrough of a solution... and creative breakthroughs - innovative ways of solving problems is kinda the point with programming.

I agree that a set of heuristics are useful... ie the equivalent to your general rules.

but AFAICT, there's no way that it can ever settle down into a set of rules in the way that laws can. Fundamentally, programming needs to remain fluid and flexible - it needs to always remain a bit of a frontier... to leave it open to new innovations. Just like art will never settle down to be "just one methodology".

What "methodology" can be put in place to always ensure that your art is exceptional?

Anything you can do reliably will rapidly cease being "exceptional."

Yes, that's true.

And in computing - anything that's not exceptional rapidly ceases to be acceptable.

I'd like to see more rationalists living together. This is something I'll keep in mind next time I move.

I'm moving into the same building as two LWers for exactly this reason.